Blockchain networks exist to provide immutable access to transaction data. It’s a fundamental requirement for any serious blockchain application to ensure that data is available whenever the user requests it. To minimize delay, Node Providers need to provide a high percentage of Uptime.

Between October 2025 and October 2026, ChainSafe's public-facing production infrastructure averaged 99.99% uptime across 14 endpoints.

What 99.99% availability means

Availability is the percentage of time a service responds correctly to requests. At 99.99%, a service can be down for about 53 minutes in a year.

While many Node Providers treat uptime as a marketing metric, this article explores the technical methodology required to sustain such reliability. We move beyond vendor-first claims to discuss the engineering reality of managing blockchain nodes, covering the strategic use of bare-metal servers and geo-distributed clusters. Whether you are managing a validator node or scaling full nodes for a high-traffic protocol, resilience is a product of rigorous automation and architectural foresight.

At a glance

• 99.99% mean availability across 14 public-facing production endpoints, trailing 12 months
• ~83 Filecoin mainnet snapshots a week
• 7 infrastructure providers across 4 regions
• 30+ networks supported since 2017

How ChainSafe built its high-performing infrastructure

Seven providers

Relying on a single cloud vendor creates a systemic point of failure. To achieve institutional-grade stability, we distribute infrastructure across seven distinct providers, including AWS, Hetzner, OVH, Scaleway, Latitude, CherryServers, and Servers.com.

By diversifying our blockchain infrastructure platform across these providers, we mitigate the risk of provider-specific outages or sudden policy shifts. This multi-layered approach ensures that our global web3 server fleet remains resilient against regional or provider-level instability, maintaining continuity for the Ethereum blockchain and other Layer 1s.

Spreading the fleet across providers limits how much of it any one provider's outage can affect.

Four regions

• Europe: Germany, Finland, France
• North America: Canada, United States
• Latin America: Brazil
• Asia-Pacific: Singapore

Geographic diversity is the second pillar of high availability.

We operate across North America, Europe, Latin America, and Asia-Pacific. This ensures that, even in the event of a catastrophic data center outage or a massive ISP disruption in one locale, traffic is automatically routed via a sophisticated content routing system.

Maintaining nodes across four continents minimizes latency and provides a hedge against regional outages that can impact your node connectivity.

What ChainSafe builds

RPC endpoints

An RPC node serves as the vital link between a dApp and the ledger. It’s the interface applications use to read data from a blockchain and send transactions back to it. For example, when an Ethereum RPC endpoint falters, the associated blockchain application effectively goes dark. 

Wallets, dapps, and indexers all depend on RPCs. When the RPC goes down, every application using it goes down too, regardless of how healthy the chain itself is.

ChainSafe manages both public and private endpoints, ensuring that developers and institutional clients have the node access required for consistent operation. Whether you are querying a smart contract or broadcasting transactions, the underlying RPC node providers must prioritize request throughput, using load balancers to distribute volume across a pool of healthy nodes.

• Public endpoints for ecosystems that want new builders to be able to start immediately
• Private endpoints for teams that need dedicated capacity not shared with public traffic

Bootstrap nodes

A bootstrap node is the first peer a new node contacts when it joins a peer-to-peer network. It gives the new node a list of known peers so it can connect to the rest of the network. Without them, a new node cannot find its peers and fails to sync with the network.

ChainSafe operates six Filecoin bootstrap nodes: three on mainnet and three on Calibnet, Filecoin's calibration testnet. Their addresses are hard-coded into three independent Filecoin implementations: Lotus, Forest, and Venus. Two of those three clients are maintained outside ChainSafe, and anyone can verify this in each client's public source code. A client team only hard-codes a bootstrap address it trusts to stay online.

ChainSafe also operates an archival bootstrap node that serves the chain's full historical state to the network. It is currently available only on Calibnet.

Snapshots and fast sync

Syncing from Genesis is a relic of early blockchain development. To maintain high availability, modern infrastructure relies on snapshots. 

A snapshot is a copy of a blockchain's state at a specific block. A new node can start from a recent snapshot instead of replaying every block since genesis, which can take days.

By providing these periodically, we allow operators to jump-start their nodes, significantly reducing the time required to regain full node synchronization after an outage or a chain migration.

Filecoin snapshots

We produce ~83 mainnet and ~82 Calibnet snapshots weekly, ensuring that Filecoin participants can sync with minimal downtime.

• One Mainnet and Calibnet snapshot is made every two hours.
• 14-day retention.
• We aim to publish a fresh snapshot within 30 minutes of a network upgrade so that operators can resync on the upgraded chain right away.
• 7 archive diff snapshots a week.

Current snapshots are listed at forest-archive.chainsafe.dev/list.

Filecoin Archive

The Filecoin Archive acts as a long-term historical ledger, ensuring that the archival nodes maintain a complete, verifiable record of state history.

Ethereum snapshots

ChainSafe provides up-to-date Ethereum state snapshots, enabling developers and node operators to initialize nodes without waiting days for block replay.

Checkpointz nodes

Checkpoint sync allows an Ethereum consensus client to start from a recent finalized state rather than syncing from genesis. ChainSafe operates Checkpointz nodes that serve these checkpoints on two networks:

• Mainnet: beaconstate-mainnet.chainsafe.io
• Hoodi: beaconstate-hoodi.chainsafe.io 

Archive nodes and data access

A full node validates transactions, but archive nodes preserve the history of every account and state change. These nodes are essential for block explorers, analytics platforms, auditors, and institutional data teams. Operating them requires significantly higher storage overhead, often necessitating the use of bare-metal servers to prevent "noisy neighbour" issues common in shared cloud environments.

ChainSafe runs:

• 4 Ethereum archive nodes: 2 on mainnet, 1 on Sepolia, 1 on Hoodi
• 2 Filecoin archival nodes

Testnets and developer tooling

Reliability in production begins with rigorous testing. ChainSafe maintains infrastructure across 10 testnets to ensure that protocol upgrades are battle-tested before they reach mainnet.

• Ethereum: Hoodi, Sepolia, Ephemery
• Filecoin: Calibnet
• Celestia: Mocha-5
• Aztec testnet
• SSV: Hoodi
• Monad: testnet2
• Canton: dev1
• WalletConnect testnet

Ephemery

ChainSafe operates validator infrastructure for Ephemery, an Ethereum testnet that resets to genesis every four weeks. Developers can test contracts, clients, and staking setups without long sync times or years of accumulated data. This is essential for testing validator software and deployment scripts without accumulating blockchain bloat.

Calibnet storage miner

ChainSafe maintains a storage miner on Calibnet. It gives the Filecoin network a stable participant for testing upgrades before they reach mainnet.

Faucets

A faucet hands out tokens so developers can deploy and test without having to buy any. Faucets are the entry point for developers. Our secure, throttled faucets ensure that developers can test functionality without risking production capital. 

ChainSafe runs two Faucets:

• Forest Explorer Faucet: FIL on mainnet and Calibnet
• Calibnet FIL & Datacap Faucet: test FIL and Datacap for Calibnet development

Aztec sequencer

ChainSafe operates a sequencer on Aztec's incentivized testnet. A sequencer orders transactions and produces blocks for a rollup. Aztec is a Layer 2, privacy-first zk-rollup on Ethereum whose smart contracts support both private and public execution. 

Network node operations

Beyond standard RPCs, we support specialized roles that ensure the healthy functioning of distributed protocols.

Celestia Fibre

Fibre is a data availability protocol for high-throughput rollups. Celestia's validators run it alongside Celestia's existing data availability layer. Rollups send data directly to validators, which store it and sign to confirm they hold it. ChainSafe runs Fibre as a Celestia validator.

drand mainnet node

ChainSafe is a long-standing member of the League of Entropy, alongside Cloudflare, Ethereum Foundation, Protocol Labs, and 13 others. ChainSafe operates a mainnet node for drand, a distributed randomness beacon that produces publicly verifiable random values at fixed intervals. drand is a key component for decentralized gaming and financial systems that require verifiable randomness.

WalletConnect node operator

The WalletConnect Network connects wallets to dapps through end-to-end encrypted sessions. ChainSafe runs a service node for the network, so wallets and applications can connect across chains without having to host their own node clusters.

SSV node operator

ChainSafe operates a node for SSV Network. SSV uses distributed validator technology (DVT), which splits Ethereum validator keys into shares held by several independent operators, so the validator keeps running if one operator goes offline. No single operator holds the full key. 

Validator operations

We operate validators across several ecosystems. A validator node is the heartbeat of any Proof of Stake chain, producing and verifying blocks. They require robust hardware, strict adherence to the consensus algorithm, and top-tier security.

ChainSafe operates validators on the following networks:

• Ethereum Validators: ChainSafe runs ~8,000 validators as a node operator in Lido's Curated Module, the set of operators approved by Lido DAO governance. We also run validators through Rocket Pool, which supports Lodestar, our consensus client, as part of its Smart Node stack.
• Canton validator: a private gateway for participants to transact and manage digital assets on Canton. It keeps the ledger synchronized in real time and validates Daml smart contract execution for the transactions it is party to.
• Celestia validator: takes part in consensus and governance.
• zkVerify validator: in the network's initial active set since genesis, contributing to block production, ZK proof verification, and consensus.

Storage and content access

Orbitor IPFS public gateway

An IPFS gateway connects the regular web to IPFS. It receives a request for content-addressed data over HTTP, retrieves that data from IPFS, and serves it back so any browser can load it.

ChainSafe operates Orbitor, a public IPFS gateway. It runs on 3 gateways and 3 IPFS nodes across Europe, Latin America, and Asia-Pacific.

Canton infrastructure and tooling

ChainSafe is a Canton Super Validator, and we help teams get onto Canton and build on it. We host the infrastructure under teams' own validators, provide development environments, support onboarding, and build enterprise-ready tooling.

Infrastructure hosting for Canton validators

This is for teams that operate their own validator. ChainSafe provides and runs the infrastructure underneath it: servers or an isolated Kubernetes environment, networking, monitoring, and backups. The team holds its own party and keys and operates the validator itself.

Devnet and testnet environments

We set up hosted or local Canton devnet and testnet environments that teams can connect to and build against.

Canton Network onboarding support

We give hands-on help getting a node running in a team's own environment. That covers deployment, networking and allowlisting, upgrades, and the operational practices we use on our own infrastructure.

Ramp-up support for our Canton products

We help teams adopt the three tools below. That can mean integrating the ERC-20/CIP-56 middleware, running an x402 facilitator on their own node, or using DAML Reason in their development workflow. 

DAML Reason

DAML Reason is a Model Context Protocol (MCP) server built for Canton development. It accelerates development workflows by providing compiler-validated patterns, ensuring that smart contracts are secure by design.

• It bases its code suggestions on 1,461 compiler-validated Daml patterns.
• It catches authorization and compliance errors before code review.
• It includes CI automations that run Daml tests, builds, and sandbox environments on every push.

ERC-20/CIP-56 middleware

ChainSafe's ERC-20/CIP-56 middleware makes Canton compatible with EVM networks. Users sign Canton transactions in MetaMask and manage CIP-56 tokens the same way they manage ERC-20s on Ethereum, with no Daml rebuild required. Each partner gets a dedicated relayer that locks assets 1:1 on Ethereum and mirrors them natively on Canton.

Canton x402 Facilitator

x402 is an open payment standard built on the HTTP 402 "Payment Required" status code. When an API endpoint requires payment, the ChainSafe x402 Facilitator settles it on Canton, so AI agents and institutional participants can pay per call without a human in the loop.

Signing is non-custodial: private keys never leave the payer. Canton facilitators don't sponsor gas so that any participant can run their own facilitator on their own node, behind their own firewall.

Security practice

Infrastructure that stays online also has to keep its keys safe. ChainSafe practices three practical approaches to cover all our bases. 

• Keystores never go into code repositories. Keys stay out of version control, so a leaked or misconfigured repository can't expose them.
• Standalone validator keys are generated air-gapped.
• Threshold keys are never assembled. Instead, they're generated as shares across our signing fleet.

Why does an R&D team run production infrastructure?

ChainSafe’s identity as both a protocol developer and an infrastructure operator provides a unique feedback loop. When we build clients like Lodestar or Forest, we are the first to test them in production. This validator node setup allows us to catch bugs, refine consensus mechanisms, and maintain client diversity—the single greatest defence against network-wide outages.

Our commitment to running production infrastructure ensures that our R&D remains grounded in the realities of maintaining 99.99% uptime.

It is ChainSafe’s unique position that allows us to fine-tune our infrastructure to ensure peak compatibility with the protocols we operate. This leads to faster diagnosis: when a node misbehaves after a network upgrade, our engineers often know the client code and the upgrade specification. They can tell whether the fault lies in the infrastructure or the protocol and act on it directly rather than escalating to another team. That shortens the time from page to fix.

Conclusion

Achieving 99.99% uptime is the result of architectural discipline. It requires the right mix of Dedicated nodes, geo-redundancy, and constant security monitoring. By understanding the underlying consensus mechanisms and maintaining a diverse set of blockchain nodes, you protect your blockchain application against the unpredictability of decentralized networks.

Treating uptime as an engineering philosophy transcends it being simply a feature or marketing metric. Whether you’re selecting RPC node providers or managing your own node operator fleet, the benefits of high availability never become a point of contention when decentralizing your infrastructure.

Talk to us about your network's infrastructure ⤵️

For further guidance on optimizing your infrastructure, please visit us at ChainSafe.io or contact us directly!

FAQ

What is the difference between a full node and an archive node?

A full node stores recent history and current state. Archive nodes store all state since genesis, which is essential for auditing and forensic data analysis.

How do we mitigate DDoS attacks?

We implement high-level security solutions, including traffic shaping, Web Application Firewalls (WAFs), and monitoring of Cloudflare Ray ID patterns to isolate malformed data or SQL injection attempts.

How do compute units impact uptime?

Effective management of compute units and API credits prevents resource exhaustion. We utilize a normalization formula to ensure that heavy requests don't degrade service for other users, maintaining stable API uptime.

Which Filecoin clients use ChainSafe's bootstrap nodes?

Lotus, Forest, and Venus all hard-code ChainSafe's bootstrap node addresses: three on mainnet and three on Calibnet. You can verify this in each client's public source code.

How often does ChainSafe produce Filecoin snapshots?

Every two hours, which is about 83 mainnet and 82 Calibnet snapshots a week, kept for 14 days.

What uptime does ChainSafe's infrastructure achieve?

99.99% mean availability across 14 public-facing production endpoints over the trailing 12 months. Six endpoints were at 100%, and the lowest was 99.96%. This is measured performance, not an SLA.

Which providers and regions does ChainSafe use?

AWS, Hetzner, OVH, Scaleway, Latitude, CherryServers, and Servers.com. The regions are Europe (Germany, Finland, France), North America (Canada, US), Latin America (Brazil), and Asia-Pacific (Singapore).

Which networks does ChainSafe support?

ChainSafe has supported 30+ networks since 2017. Current operations include Ethereum, Filecoin, IPFS, Canton, Celestia, drand, WalletConnect, SSV, and zkVerify, as well as 10 live testnets.

About ChainSafe

ChainSafe is a leading blockchain research and development firm specializing in protocol engineering, infrastructure development & operations, and co-development.

ChainSafe creates solutions for developers and teams across web3. As part of our mission to build accessible, improved tooling for developers, ChainSafe embodies an open-source, community-guided ethos to advance the future of the internet.

Website | Youtube | Twitter | Linkedin | GitHub